Staging Problems NW 7.2.1

[yadnakis]

Hi

I have Networker 7.2.1 in an AIX server.
I configured staging from disc to tape. But i have one problem. If there is a corrup ssid in the disk, the stage proccess try to save it to the tape, but it can´t do it completelly . Legato try to save it again, but can`t do it in the same tape (cause the tape has the corrupt ssid save it), and pick another one, with the same results .
Legato try to save the corrupt ssid, until i stop the nsrstage proccess.
The corrupt ssid has been save 7 times, in diferent tapes

I`ts a known bug of legato?

How could i resolve it?

Thanks in advance

 

[605]

This is the natural behavior. A SSID can only exist once on each tape. Consequently, if you rewrite the same save set,
it will get a new SSID assigned. However, if you deal with an existing save set, this is not possible any more.

The SSID is written at the beginning of the save set. That allows NetWorker to retrieve the information when scanning the media. If the save set will be aborted later for whatever reason, NW will reject a second write attempt to the same tape again because this will conflict with the rule mentioned above.

Obviously the automatic staging process retries until each appendable media in the same pool has been used. This of course could be improved but i do not see any way how you can change the behavior with the current version. NW obviously assumes that the disk will not contain a corrupt save set.


To overcome the problem you can
- manually stage the save set and delete the save set from all other media
- delete the save set from the disk (if allowed)

 

[yadnakis]

Hi

Thanks for your reply.

I had make a manual script with nsrstage command.

The behavior is diferent in later versions?

Thanks

 

[605]

I do not know whether the failure behavior has changed with a later version? Unless you manipulate the files directly it is hard to create bad save sets on a file device.

However, i am sure that the rule not to allow the same save set to be stored on the media twice, is fundamental and will never change.

 

[Tonduma]

605, in 7.3 NW for AIX version occur the same and it's necessary to do manual staging ?

 

[605]

If NW can detect a "defective" save set (for example an aborted) one, it will delete it, if possible.

With respect to when this will happen, there are slight differences between file and advanced file type device. They are listed in the docs.

However, i am sure that there will still be a problem if the saved file will somehow corrupted on disk.